LiteFold/UniProtKB

UniProtKB Processed 是一个基于UniProt知识库（UniProtKB）的蛋白质序列数据集，旨在为用户提供全面、高质量且免费访问的蛋白质序列数据，这些数据带有功能注释信息。数据集包括三个主要源集：Swiss-Prot已审核的规范蛋白质（574,627条记录）、Swiss-Prot替代异构体序列（41,333条记录）以及TrEMBL未审核的蛋白质（202,556,314条记录），总计203,172,274条蛋白质记录。数据集还包含总残基数75,747,523,712、序列分片205个、蛋白质条目表分片615个等附加指标。它支持特征提取任务，并提供了训练、验证和测试分割，适用于生物信息学和蛋白质研究。数据以JSONL格式存储，并包含元数据记录和序列文件。

UniProtKB Processed is a protein sequence dataset based on the UniProt Knowledgebase (UniProtKB), aiming to provide users with a comprehensive, high-quality, and freely accessible set of protein sequences annotated with functional information. The dataset includes three main source sets: Swiss-Prot reviewed canonical proteins (574,627 records), Swiss-Prot alternative isoform sequences (41,333 records), and TrEMBL unreviewed proteins (202,556,314 records), totaling 203,172,274 protein records. Additional metrics include total residues of 75,747,523,712, sequence shards of 205, and protein-entry table shards of 615. It is designed for feature-extraction tasks and provides train-validation-test splits, suitable for bioinformatics and protein research. Data is stored in JSONL format and includes metadata records and sequence files.